3 ms·
Yeah I can show you one of the systems: GC of death: https://imgur.com/a/6RAX7 https://imgur.com/a/6RAX7 Write latency: https://imgur.com/a/Us0MO https://im
by StreamBright 9y ago
Yeah I can show you one of the systems:
GC of death:
https://imgur.com/a/6RAX7 https://imgur.com/a/6RAX7
Write latency:
https://imgur.com/a/Us0MO https://imgur.com/a/Us0MO
Typical Cassandra user has these sort of problems. I have one client where they identified the issue (data modeling) and fixed it by redesigning their tables before I got there. Some of the Cassandra users are not even aware. The company where the pictures are from engaged me because the system did not meet with business requirements anymore. Could not insert data into the cluster, their ETL jobs were running for 20 hours and if one failed they did not have data for their business. We could speed it up to run it for 4 hours without remodeling the data. With remodeling it Cassandra was not the bottleneck anymore.
- riku_iki 9y agoThese graphs are meaningless in a vacuum, you don't know how would behave alternative Go based solution on similar hardware under the similar load.
- StreamBright 9y agoSure, this is why there is a production system surrounding them. I am not really interested in theoretical possibilities just observing reality here.
- riku_iki 9y agoWhat I am saying is that you are observing one side of reality. It is very possible that even with this kind of graphs, Cassandra is capable to handle much larger traffic then custom Go solution. Cassandra is usually competitive in open benchmarks.